Interfacial tension of oil/water emulsions with mixed non-ionic surfactants: comparison between experiments and molecular simulations

Abstract

Stable oil/water emulsions are usually obtained by using mixtures of different surfactants. Such systems display synergistic interface stabilizing effects, which have not been fully elucidated yet. Moreover, in many applications surfactants are added at concentrations well above their critical micellar concentration (CMC), and this regime has not been thoroughly explored in the literature as well. Here, we investigate oil/water emulsions through oil/water interfacial tension using two common non-ionic surfactants, Tween 80 and Span 20, in the concentration range C (0.3–1 wt%) well above their respective CMCs. Mesoscale molecular simulations coupled interfacial tensiometry experiments to characterise these interfaces at a molecular level. Interfacial tension γ was measured by a pendant drop technique. Coarse-grained calculations provided a microscopic view of the interface at the molecular level (i.e. surfactant arrangement, interface thickness), and were employed to extend the study to those surfactant concentrations where experiments could hardly provide reliable data, if any. We found a significant synergistic effect between Tween 80 and Span 20, with low molecular weight Span molecules occupying free spaces between the much larger, bulky Tween compounds. The surfactant intermolecular interactions could be associated to a striking decrease of interfacial tension in going from pure surfactants to a mixture at the same total weight concentration. Furthermore, the interface was found to exhibit a spatial inhomogeneity with a “patch-like” organisation, reminiscent of microphase separation. Our results show that the proposed, combined experimental/in silico approach provides relevant insights for several industrial applications, such as emulsion stability and oil spill remediation.

---

Introduction

Emulsions are thermodynamically unstable systems of two immiscible liquids (e.g. oil and water), where one phase (the droplet or disperse phase) is embedded in another one (i.e. the continuous phase). In order to obtain long-term stable emulsions, a surfactant is usually added to the system to negatively interfere with the phase separation process.¹ The surfactant molecules, by virtue of their amphiphilic nature, migrate toward the oil–water interface whereby, by inhibiting coalescence^2,3 and decreasing interfacial tension, they exert their role as interface stabilisers. Throughout this process, interfacial tension is a critical parameter which, as reported since Traube's⁴ work, is strictly connected with the surfactant molecular structure; indeed, amphiphilic molecules featuring longer hydrocarbon tails in their hydrophobic portions endow the resulting emulsions with lower interfacial tensions. Moreover, if the surfactant concentration exceeds its corresponding critical micellar concentration (CMC) value, the surfactant molecules spontaneously self-assembly into micelles, thus plummeting the interfacial tension to a minimum.

Systems capable of achieving low oil–water interfacial tension are of great importance in many industrial fields, such as enhanced oil recovery⁵ and nanomaterial synthesis.⁶ Although the measurement of ultralow interfacial tension is therefore a relevant issue in these fields, the spinning drop method and a recently developed microfluidic method⁷ are currently the only reliable experimental techniques capable to fulfill this need. The interfacial activity of a surfactant originates from its interfacial efficiency and adsorption density, both of which are affected by the molecular structure and the assembly behaviour at the interface, and are related to the interactions between oil, water, and surfactants molecules. Thus, the investigation of the interfacial molecular behaviour of surfactants, co-surfactants or other molecules is not only important for a complete understanding of the mechanism of oil–water interfacial tension, but is also essential in efficiently identifying application technologies.

Among the class of non-ionic surfactants, the CiEj⁸ (alkyl polyglycolethers, also known as CiEOj, or Brij) family and the Tween (polysorbate) and Span (sorbitan) series are the most exploited in several applications. For example, Tween 80 (polyoxyethylene (20) sorbitan monooleate–hydrophilic surfactant) and Span 20 (sorbitan monolaurate–hydrophobic surfactant) are ubiquitous in petroleum,⁹ food,^10–13 drug delivery¹⁴ and cosmetics industry.¹⁵ Tween 80 has also been tested in aqueous solution for volatile organic compounds (VOCs) abatement,¹⁶ for aromatic hydrocarbons removal from contaminated soils¹⁷ and oil spill remediation as reported by Huston and Larson,¹⁸ Kirby et al.¹⁹ and Athas et al.²⁰ (see also reference therein). Non-ionic surfactants are also widely used to produce nanoemulsions.⁶ It is well known that a combination of two surfactants leads to lower interfacial tension (which, in turn, corresponds to lower CMC values) compared to the equivalent emulsion containing one type of surfactant only, ultimately allowing the obtainment of very stable systems.¹⁵ Thus, some “synergism” between the two amphiphilic molecules is expected to come into play by enhancing the mutual compatibility between water and oil.^6,21 This mechanism is intuitive for a combination of ionic/non-ionic surfactant, due to Coulomb, hydrogen bonding and ion-dipole interactions. Nevertheless, a similar effect is reported for non-ionic/non-ionic surfactant mixture, where, despite the existence of weaker molecular interactions, a higher mutual solubility is found as compared to the single surfactant case.²² In 1972, Boyd et al., investigated Tween and Span blends and found a direct proportionality between emulsion interface and stability: the more compact its interface, the more stable the emulsion.²³ According to these authors, two main molecular aspects contribute to this phenomenon: the geometrical packing of the surfactant heads and the length of their hydrophobic portions. The first concept can be explained by considering that a hydrophilic surfactant with a bulky head can be coupled with a hydrophobic one that usually has a smaller head group, the latter filling the void spaces between the large head groups of the former. The role of surfactant tail length involves van der Waals interactions between their hydrocarbon chains; in particular, the longer the tails (at least eight carbon atoms), the stronger the intermolecular hydrophobic interaction forces, and, eventually, the more stable the corresponding interfacial film. Moreover, in the case of non-ionic/non-ionic surfactants, a further, favourable contribution to synergism occurs when the length of the hydrocarbon tails of the two surfactant is similar.²⁴ The seminal arguments of Boyd et al. have been confirmed later by the work of Brochu et al., who have demonstrated the efficiency of Span–Tween blends as oil dispersants.⁹

Due to the great interest in emulsion development, a lot of works refers to the interfacial tension characterisation, which is one of the most important parameters related to system stability. Such interfacial properties have been thoroughly investigated in the dilute regime, i.e. nearby or below the CMC, for a single surfactant system. Within the same regime, some insights can be found also for the mixed surfactants case.^1,24,25 However, the high surfactant concentration range (≫CMC), which is still largely unexplored in the literature, is quite relevant for industrial applications.⁶ Unfortunately, the classical techniques used for the measurement of the interfacial tension, such as the pendant drop analysis, are not quite suitable to characterise highly concentrated surfactant systems due to the viscoelastic properties of the interfacial layer. Bulk mass transfer of the adsorbed surfactants film in combination with high interfacial shear elasticity, could induce shrinkage of the pendant drop and give rise to buckling or crumpling phenomena at the surface.²⁶

To overcome experimental limitations, numerical simulations offer a viable alternative to investigate interfacial properties of emulsions having ultralow interfacial tension or high surfactant concentration.²⁷ An overview the molecular modeling tools used to predict surfactant properties at the interface and CMC has been reported recently by Creton et al.²⁸

Despite accurate predictions, most atomistic computational techniques like molecular dynamics (MD) simulations are still computationally expensive when one wants to reach length and time scales comparable to those of experiments.

Coarse-grained (CG) simulations provide a bridge between the atomistic and macroscopic worlds. The CG simulation approach considers clusters of atoms or molecules as single particles, called beads. Compared to MD, CG models reduce the number of degrees of freedom, and hence the corresponding computational time, by lumping the detailed chemical forces into the form of interaction parameters. Among all CG methods, Dissipative Particle Dynamics (DPD)^29–31 is a particle-based mesoscopic technique, which involves a factor of 1000 time steps increase and a factor of 10–100 size-scale increase with respect to atomistic modeling. As such, it holds a great momentum in the prediction of mesoscale property and can be considered as an additional, useful tool in chemical process applications.

Many achievements have been reported in the use of DPD for liquid mixtures and complex systems.^32–37 E.g., recently, Boromand et al. applied DPD method to estimate shear viscosity of Newtonian liquids.³⁸ Specifically in the field of the present paper, Rekvig et al. used the DPD method to simulate and predict surfactant behaviour at the interface in a single surfactant system.^4,29 The surfactant efficiency was found to depend both on chain length, thereby confirming previous literature evidences, and on branching. The use of branched surfactants has a positive effect on interfacial tension when the hydrophilicity of the head groups overcomes the steric repulsion between the tails, since such condition leads to the formation of a compact layer at the interface between oil and water. The combination of DPD with Monte Carlo calculations was exploited by the same author to provide an estimate of the surfactant bulk concentration.²⁹ Ginzburg et al. used DPD and self-consistent field theory (SCFT)³⁰ to model interfacial tension in an emulsion system containing a single non-ionic surfactant (of the CiEOj family) and a comparison with experimental data from Rosen and Murphy³⁹ was given. Quite good agreement between experiments and simulations was found even if, both SCFT and DPD did not properly take into account micelles formation in the bulk phase, a key-issue recently faced by Yang and Sun⁴⁰ in their coarse-grained simulations. Tang et al. recently reported about the property of Tween 80 at the oil–water interface by MD simulations.³³ Notably, however, most of numerical simulations are focused on a single surfactant system, whereas the characterisation of mixed surfactant emulsions, that are ubiquitous in industrial applications is still lacking, even if a recent paper regarding oil–water interfaces characterised by a mixed layer of Tween 80 and polyethylene glycol has been recently published.¹⁸

Given this scenario, in this work we investigate the oil/water interfacial tension under static conditions when two widely used surfactants, Tween 80 and Span 20, in different concentrations are exploited. Static interfacial tension measurements are obtained by pendant drop tensiometry by using a mineral oil, which is widely used in a number of applications and products. The experimental data are compared with results obtained by mesoscale simulations on dodecane, taken as prototypical representative of the mineral oil used in the experiments, and two model molecules of Tween 80 and Span 20. This choice is justified by the polydispersity of both mineral oil and surfactants, which we deemed unnecessary to include in the simulations being outside the scope of this work. Hence, representative model molecules were used in the calculations, as it is commonly done in molecular simulations.¹⁸

At last, the mesoscale simulations allow predicting interfacial tension in a surfactant concentration range not accessible with the pendant drop method, concomitantly yielding insight on surfactants distribution and partitioning at the interface.

Results and discussion

Experimental results

We measured static interfacial tension γ for the different surfactant systems. Table 1 reports the mean values of the static interfacial tension γ for the different surfactant systems considered. Each value represents an average over the analysis of at least five drops obtained by the pendant drop technique. Interfacial tension values in the case of pure interfaces – oil-in-water or water-in-oil systems – were found to be about 20–25 mN m⁻¹ (Fig. 1a and c). A slight difference between the two configurations can be attributed to the polydispersity of the mineral oil and possible impurities present therein. In the systems containing mixture of surfactants, the ratio between Span 20 and Tween 80 was kept constant and equal to 1, as this value was found to provide a better emulsion stability compared to different surfactants ratios. Surfactant concentration in both phases was varied between 0.3% to 1% wt, a concentration range well above the CMCs of both surfactants (2.3 × 10⁻⁵ M for Span 20 (ref. 41) and 1.2 × 10⁻⁵ M for Tween 80 (ref. 17), respectively). Tween 80 and Span 20 were dissolved separetely into the water and oil phase respectively.

Table 1 Interfacial tension measurements performed by pendant drop technique for different systems

Continuous matrix	Drop phase	γ^a [mN m^−1]
a Mean values, standard dev. of ±0.2 mN m^−1.
Pure mineral oil	Deionized water	20.5
Oil + 0.3% wt Span 20	Water + 0.3% wt Tween 80	2.1
Oil + 0.5% wt Span 20	Water + 0.5% wt Tween 80	2
Oil + 0.7% wt Span 20	Water + 0.7% wt Tween 80	2
Oil + 1% wt Span 20	Water + 1% wt Tween 80	1.8
Deionized water	Pure mineral oil	24.5
Water + 0.3% wt Tween 80	Oil + 0.3% wt Span 20	2.4
Water + 0.5% wt Tween 80	Oil + 0.5% wt Span 20	2.1
Water + 0.7% wt Tween 80	Oil + 0.7% wt Span 20	1.8
Water + 1% wt Tween 80	Oil + 1% wt Span 20	1.7

---

Fig. 1 Frames analysed for the following systems: (a) water in mineral oil (pendant drop); (b) water + 0.5% wt Tween 80 in mineral oil with 0.5% wt Span 20; (c) pure mineral oil in pure water (rising drop). Needle external diameter is 1.57 mm.

In such high concentration range a nearly constant value of the interfacial tension (equal to the one at the CMC) is expected. Furthermore, under these conditions, no dynamical adsorption has been observed, i.e. γ reached the equilibrium value in a very short time scale. All analysed drops were observed for extended time periods (about 30 min) and no interfacial tension changes during this time scale were detected, so that any depletion phenomena could be ruled out.

In the water drop case (i.e. drop pending from the needle), we detected a value of about 2 mN m⁻¹ with a slightly decreasing trend (albeit within data standard deviation). Reverting the pendant drop configuration from water drop to oil drop (i.e. oil drop rising from a hook-shaped needle – Fig. 1b) confirmed both the interfacial tension values and the relevant slightly decreasing trend although, in this case, the difference between the measured values was above the standard deviation.

At the highest surfactant concentrations considered (>1% wt), or when dissolving the two surfactants in the same phase at any concentration value between 0.3% and 1% wt, the pendant drop method failed, as shown in Fig. 2, since the formation of a filament, instead of a drop, occurred. This is likely due to the viscoelastic properties of the interfacial surfactant layer. Accordingly, no measurements of interfacial tension were carried out under these conditions nor further investigation on system viscoelasticity was undertaken as this issue was outside the scope of the present work. The interested reader is referred to the literature^42,43 for further insight on the viscoelastic behaviour of these systems.

Fig. 2 98% of mineral oil + 1% wt Span 20 + 1% wt Tween 80 (3 × 10⁻² M) in 99.5% of water + 0.5% wt Tween 80 (3 × 10⁻³ M); example of filament formation.

Computational results

We started our analysis by investigating the characteristics of the behaviour of each single surfactant (1% wt) at the oil–water interface. Simulation box cross-sections with equilibrium nanoscale structures are given in Fig. 3.

Fig. 3 DPD simulation snapshots of water/dodecane/Tween 80 (1% wt) (a and b) and water/dodecane/Span 20 (1% wt) (c and d) systems. Surfactant molecules are depicted in stick style and a transparent grey and brown field is used to represent water and oil, respectively. (c) and (d) show a zoomed view of the interface at the equilibrium for each system, where the hydrophilic surfactant segments are colored in steel blue and the hydrophobic ones in sea green.

The simulation snapshots revealed that that the surfactants are located at the interface, with the head groups in the water phase and the tail groups in the oil phase. Moreover, micelle formation was detected in both cases as expected, since surfactant concentrations were above the corresponding CMCs.

As evident from Fig. 3, surfactant molecules are subjected to two competing factors: on one hand, they show a strong tendency to adsorb at the interface but, at the same time, since the oil–water surface has been saturated, for some surfactants aggregation into micelles becomes the most favorable process. It should be noted that we do not see any Span reverse micelle in the oil phase. We surmise that this could be attributed to the small amount of oil present. However, it is also known that when Tween micelles are present in the aqueous phase, Span molecules are easily dragged from the oil phase into the aqueous one thus forming mixed micelles.⁴⁴

In order to gain insights into the characteristics of the interfacial performance of Tween 80 and Span 20, the interfacial phases in Fig. 3b–d were analysed in details.

The normalised density profiles associated with a set of the relevant chemical specie in the direction perpendicular to the monolayer are reported in Fig. 4.

Fig. 4 Normalised density profiles along the direction perpendicular to the water/oil interface for the nonionic surfactants Tween 80 (1% wt) (left) and Span 20 (1% wt) (right) systems. Color legend: water, blue; oil, grey; head group, green; tail group, orange.

For the surfactant, the plots correspond to averages taken over sites lying at the two main components of the molecule, i.e. the hydrophobic tail and the complementary hydrophilic section.

In the case of Span 20, both profiles followed a Gaussian-like distribution, as seen in the right panel of Fig. 4 whilst, when Tween 80 is considered, the hydrophilic segment is found to penetrate deeper into the underlying substrate (Fig. 4, left panel). This can be attributed to the bulkier nature of Tween 80 head group.

We could gauge the previous observations on more quantitative ground by evaluating additional practical parameters related to the former density distributions. Among them, interfacial thickness is an important physical parameter that provides a quantitative measure of the size of the interface. Accordingly, we computed the interface thickness from the region along which the local density departed from the bulk value in the 90–10% interval. The corresponding values of this quantity were found to be 38 Å and 29 Å for Tween 80 and Span 20, respectively. Another key parameter characterising these systems is the molecular area for surfactant at the interface (A), for which we found the values of 66 ± 4 Ų and 32 ± 5 Ų for Tween 80 and Span 20, respectively, in agreement with the corresponding estimations available in literature.^45,46

Next, we determined the value of the relevant interfacial tension γ from the difference in normal and tangential stress across the interface.⁴⁷ If z is the direction perpendicular to the surfactant monolayer, γ can be derived using the Irving–Kirkwood equation⁴⁷ by integrating the stress difference over z (eqn (1))

in which p_xx, p_yy, and p_zz are the three diagonal components of the pressure tensor p. At the concentration considered, Tween 80 was able to reduce the water–oil interfacial tension from 48 ± 4 mN m⁻¹ (pure water–oil emulsion) to 10 ± 3 mN m⁻¹. Contextually, Span 20 was found to be slightly less efficient in this respect, since the final value of γ was found equal to 16 ± 4 mN m⁻¹. Notably, the predicted γ values are in agreement with those reported in literature,^41,48 ultimately validating the DPD models of the surfactant systems used in this work.

Next, employing the above-validated surfactant models, several mixtures of Tween 80 and Span 20 were simulated at the mesoscale level. Fig. 5 shows a view of the mixed systems at 0.3% wt and 1% wt surfactant concentration, respectively.

Fig. 5 DPD simulation snapshots of two water/oil/surfactant systems with different surfactant concentration: 0.3% wt (a) and 1% wt (b). Tween 80 molecules are depicted in stick style, whilst Span 20 molecules are portrayed as sticks-and-balls with hydrophilic segments colored in steel blue and hydrophobic colored in light sea green. A transparent grey and brown film is used to represent water and oil, respectively. (c) Planar top view of the simulated oil/water interface in the presence of a Tween 80/Span 20 mixture at 1% wt concentration. Tween 80 molecules are depicted in steel blue stick style, whilst Span 20 molecules are portrayed as orange sticks-and-balls.

A significant parameter in surfactant mixtures is the molecular interaction between two surfactants, normally denoted as β. β expresses the molecular interaction between two different surfactants in a solution in relation to the self-interaction of the individual surfactants prior to mixing (eqn (2))^1,49

where W₁₂ is the molar interaction energy between the mixed surfactants, W₁₁ is the molar self-interaction energy of the first surfactant, W₂₂ is the molar self-interaction energy of the second surfactant, R is the molar gas constant, and T is the absolute temperature.

Thus, the sign of β tells about the nature of interaction while the magnitude indicates its strength: the greater the value of β, the stronger the surfactant interaction. Negative values correspond to synergic action of the components (more attractive or less repulsive than the self-interaction of the surfactants individually), while positive values mean antagonism between the two molecules (more repulsive or less attractive than the self-interaction of the surfactants individually). Values closed to zero indicate ideal mixing.

Our calculations predicted a value of β of −6.2, evidencing a significant synergistic effect between Tween 80 and Span 20. There are two main factors that govern the interaction between non-ionic groups. These are steric repulsions and electrostatic interactions, both of which depend on the chemistry and size of the head and the tail group. After mixing, our simulations revealed that the low molecular weight Span molecules occupy the free spaces between the huge, bulky Tween molecules, thereby creating a condensed layer at the interface characterised by a negative value of β. The bulkiness of the Tween 80 hydrophilic groups decreases the self-repulsion in nonionic surfactant featuring a linear hydrophilic group. At the same time, attractive non-bonded interactions between neighboring hydrophobic chains in mixed surfactants produce stronger surfactant molecular interactions, resulting in a more negative β parameter. Indeed, the interaction parameter a_TS-TT = 29 is less repulsive than the corresponding self-interaction counterparts a_TT-TT = 35, a_TS-TS = 33.

Our simulations revealed that, regardless of surfactant concentration, the interface exhibits some inhomogeneity in the direction parallel to the interface. Indeed, the top view of the interface shown in Fig. 5c reveals a “patch-like” organisation: Tween 80 has a large hydrophilic cross-sectional area relative to its tail group, so molecules with small hydrophilic cross-sectional area like Span 20 complement molecules in Tween 80 films. Accordingly, the adsorption of a mixture of surfactants is likely to bring more complexity in the resulting interface; this, in turn, results in high efficiency in decreasing the oil–water interfacial tension (first and second column of Table 2).

Table 2 Experimental γ_exp and simulated γ_sim surface tension of different binary mixtures of Tween 80 and Span 20, area per surfactant A and relative interface thickness δ as derived from mesoscale calculation

Composition [% wt]	γexp [mN m^−1]	γsim [mN m^−1]	A [Å^2]	δ [Å]
a Experimental values could not be obtained at this concentration due to the formation of a viscoelastic filament (see subsection on Experimental results).
Oil–water	20.5 ± 0.2	48 ± 4.0	—	—
1% Tween 80	n.a.^a	10 ± 3	66 ± 4	38
1% Span 20	n.a.^a	16 ± 4	32 ± 2	29
0.3% Tween 80/0.3% Span 20	2.1 ± 0.2	0.9 ± 0.3	40 ± 2	46
0.5% Tween 80/0.5% Span 20	2.0 ± 0.2	0.8 ± 0.2	41 ± 3	43
0.7% Tween 80/0.7% Span 20	2.0 ± 0.2	1.1 ± 0.2	40 ± 3	42
1% Tween 80/1% Span 20	1.8 ± 0.2	0.7 ± 0.3	44 ± 2	48

---

These synergistic effects were also confirmed by the average molecular area available for each surfactant at the interface A (third column of Table 2). The ideal mixing value A_ideal = X₁A₁ + (1 − X₁)A₂ (where X₁ refers to the molar fraction of the component 1 in the monolayer, and A₁ and A₂ are the value of the cross-area before mixing), was estimated to be equal to 46 Ų on average, with a mean value of X₁ = 0.4 as derived from the equilibrium interface for Tween 80. This value is larger than the corresponding calculated mean value A, clearly indicating contraction upon mixing, higher degree of packing and effectiveness in adsorption at the interface.

To analyse the equilibrium behaviour of Tween 80/Span 20 in the dodecane/water system, the density profiles normal to the interface were calculated and are shown in Fig. 6 for the systems at 0.3% wt and 1% wt surfactant concentration, respectively, as proof-of-concept. As it is clear from these graphs, the surfactants molecules are well distributed along the water–oil interface, resulting in a stable monolayer. A closer look revealed that, in presence of a surfactant mixture, the width of the interfacial region increases with respect to the pure component cases, being approximately equal to 45 Å on average (fourth column of Table 2).

Fig. 6 Normalised density profiles along the direction perpendicular to the water/oil interface for the nonionic surfactants mixtures at 0.3% (a) and 1% (b) concentration. Color legend: water, blue; oil, grey; head group, green; tail group, orange; full symbols, Tween 80; open symbols, Span 20.

When Tween 80 and Span 20 were used together, the ethoxylated parts of Tween molecules were found squeezed into the subphase, the relevant density profiles being wider and shifted to the right. Concomitantly, those parts of the Tween hydrocarbon chains located in the oil phase penetrated between the adsorbed Span molecules, in agreement with what recently reported by Tang et al.³² for Tween 80 water/alkane interfaces. This reduced the distance between adjacent hydrocarbon chains, thereby increasing the probability of attraction between the surfactant molecules. Moreover, we observed extensive water permeation of the hydrophilic head groups which ultimately gave rise to a thick hydrated ethoxylate layer.

The discussion reported above and values in Table 2 highlight that the interfacial properties of the systems under investigation do not depend on Tween/Span concentration, since, according to the simulations, the interface quickly became saturated and excessive surfactant molecules self-assembled into micelles rather than further overcrowding the interface, in reasonable consistence with the corresponding experimental evidences. At the same time, some differences between predicted and simulated interfacial tension were to be expected, which could be sensibly ascribed to three main reasons: (i) the fact that interfacial tensions were experimentally measured for mixtures of non-monodispersed surfactants while we assumed to model each surfactant as monodispersed, (ii) the light mineral oil used in the experiment as opposed to the pure dodecane phase considered in the simulations, (iii) the uncertainty intrinsic to any calculation.⁵⁰

Conclusions

The oil–water interface is the most relevant feature governing the chemico-physical behaviour of emulsions. In practical applications, where surfactant mixtures are used to improve emulsion stability, emulsion droplets are endowed with complex interfaces, whose structure has not been fully understood. In this study we reported novel results about emulsion interfacial properties for the popular pair of non-ionic surfactants Tween 80 and Span 20 in the high concentration regime. The experimental results obtained by the pendant drop method have been compared to mesoscale molecular simulations with good agreement in terms of interfacial tension. Furthermore, through simulations, we provided a molecular picture of the surfactant partitioning between interface and bulk and of their interfacial arrangement. We found that an hydrophilic surfactant with a bulky head and long hydrocarbon tail (Tween 80) coupled with a tiny hydrophobic one (Span 20) generated a densely packed interface which, in turn, resulted in significantly lower interfacial tension values as compared to those obtained for single surfactant systems (synergistic effect). In addition, we showed how mesoscale molecular simulations, coupled with experimental tensiometry, could constitute a unique tool for the investigation of systems in the high surfactant concentration regimes (≫CMC). These results can provide some fundamental insights to industrially and environmental relevant problems, such as emulsion stability and oil spill remediation.

Experimental

Materials

The transparent light mineral oil, Drakeol7®, was purchased by Penreco. Polyoxyethylene sorbitan monooleate, C₆₄H₁₂₄O₂₆, (density: 1.06–1.09 g ml⁻¹, molecular weight: 1310 g mol⁻¹, Tween 80) and sorbitan monolaurate, C₁₈H₃₄O₆, (density: 1.032 g ml⁻¹, molecular weight: 346.46 g mol⁻¹, Span 20) were purchased from Sigma Aldrich. Deionized water from Millipore was employed throughout the work.

Computational methods

DPD simulations were performed in the framework of a multiscale modeling strategy developed by our group.^51​–53 This approach is based on the systematic elimination of computationally expensive degrees of freedom while retaining implicitly their influence on the remaining degrees freedom in the mesoscopic model. Accordingly, using the information obtained from atomistic molecular dynamics simulations, we parametrized the coarse-grained (e.g., DPD) models that incorporate all essential physics/phenomena observed at the finer level.

A full description of experimental methods, DPD theory, and computational models is provided in the ESI.†

Acknowledgements

This study is related to the activities of the European network actions COST MP1106 “Smart and green interfaces—from single bubbles and drops to industrial, environmental, and biomedical applications”, COST CM1101 “Colloidal aspects of nanoscience for innovative processes and materials” and COST MP1305 “Flowing Matter”.

References

1. M. J. Rosen and J. T. Kunjappu, Surfactants and interfacial phenomena, John Wiley & Sons, 2012.
2. S. Guido and M. Simeone, J. Fluid Mech., 1998, 357, 1–20.
3. G. Marrucci, Chem. Eng. Sci., 1969, 24, 975–985.
4. L. Rekvig, M. Kranenburg, B. Hafskjold and B. Smit, Europhys. Lett., 2003, 63, 902.
5. M. J. Rosen, H. Wang, P. Shen and Y. Zhu, Langmuir, 2005, 21, 3749–3756.
6. A. Perazzo, V. Preziosi and S. Guido, Adv. Colloid Interface Sci., 2015, 222, 581–599.
7. S. S. H. Tsai, J. S. Wexler, J. Wan and H. A. Stone, Lab Chip, 2013, 13, 119–125.
8. P. Ramírez, L. M. Pérez, L. A. Trujillo, M. Ruiz, J. Muñoz and R. Miller, Colloids Surf., A, 2011, 375, 130–135.
9. C. Brochu, E. Pelletier, G. Caron and J. E. Desnoyers, Oil Chem. Pollut., 1987, 3, 257–279.
10. L. C. B. Züge, C. W. I. Haminiuk, G. M. Maciel, J. L. M. Silveira and A. de Paula Scheer, J. Food Eng., 2013, 116, 72–77.
11. Z. Chao, M. Yue, Z. Xiaoyan and M. Dan, J. Food Sci., 2010, 75, C493–C497.
12. H. D. Goff, Int. Dairy J., 1997, 7, 363–373.
13. I. Kralova and J. Sjöblom, J. Dispersion Sci. Technol., 2009, 30, 1363–1383.
14. A. López, F. Llinares, C. Cortell and M. Herráez, Int. J. Pharm., 2000, 202, 133–140.
15. C. F. Zanatta, V. Ugartondo, M. Mitjans, P. A. Rocha-Filho and M. P. Vinardell, Food Chem. Toxicol., 2008, 46, 2776–2781.
16. A. Erto and A. Lancia, Chem. Eng. J., 2012, 187, 166–171.
17. I. T. Yeom, M. M. Ghosh, C. D. Cox and K. G. Robinson, Environ. Sci. Technol., 1995, 29, 3015–3021.
18. K. J. Huston and R. G. Larson, Langmuir, 2015, 31, 7503–7511.
19. S. M. Kirby, S. L. Anna and L. M. Walker, Langmuir, 2015, 31, 4063–4071.
20. J. C. Athas, K. Jun, C. McCafferty, O. Owoseni, V. T. John and S. R. Raghavan, Langmuir, 2014, 30, 9285–9294.
21. D. Blankschtein, A. Shiloach and N. Zoeller, Curr. Opin. Colloid Interface Sci., 1997, 2, 294–300.
22. P. D. T. Huibers and D. O. Shah, Langmuir, 1997, 13, 5762–5765.
23. J. V. Boyd, C. Parkinson and P. Sherman, J. Colloid Interface Sci., 1972, 41, 359–370.
24. M. Abe, Mixed surfactant systems, CRC Press, 2004.
25. V. B. Fainerman, R. Miller and E. V. Aksenenko, Simple model for prediction of surface tension of mixed surfactant solutions, Adv. Colloid Interface Sci., 2002, 96(1), 339–359.
26. P. Erni, H. A. Jerri, K. Wong and A. Parker, Soft Matter, 2012, 8, 6958–6967.
27. S. Razavi, J. Koplik and I. Kretzschmar, Langmuir, 2014, 30, 11272–11283.
28. B. Creton, C. Nieto-Draghi and N. Pannacci, Oil Gas Sci. Technol., 2012, 67, 969–982.
29. L. Rekvig, M. Kranenburg, J. Vreede, B. Hafskjold and B. Smit, Langmuir, 2003, 19, 8195–8205.
30. V. V. Ginzburg, K. Chang, P. K. Jog, A. B. Argenton and L. Rakesh, J. Phys. Chem. B, 2011, 115, 4654–4661.
31. P. Posocco, Z. Posel, M. Fermeglia, M. Lísal and S. Pricl, J. Mater. Chem., 2010, 20, 10511–10520.
32. L. Gao, J. Shillcock and R. Lipowsky, J. Chem. Phys., 2007, 126, 015101.
33. X. Tang, K. J. Huston and R. G. Larson, J. Phys. Chem. B, 2014, 118, 12907–12918.
34. S. Jamali, A. Boromand, N. Wagner and J. Maia, J. Rheol., 2015, 59, 1377–1395.
35. D. Pan, N. Phan-Thien and B. C. Khoo, J. Non-Newtonian Fluid Mech., 2014, 212, 63–72.
36. M. L. Klein and W. Shinoda, Science, 2008, 321, 798–800.
37. J. C. Palmer and P. G. Debenedetti, AIChE J., 2015, 61, 370–383.
38. A. Boromand, S. Jamali and J. M. Maia, Comput. Phys. Commun., 2015, 196, 149–160.
39. M. J. Rosen and D. S. Murphy, Langmuir, 1991, 7, 2630–2635.
40. C. Yang and H. Sun, J. Phys. Chem. B, 2014, 118, 10695–10703.
41. L. Peltonen, J. Hirvonen and J. Yliruusi, J. Colloid Interface Sci., 2001, 240, 272–276.
42. S. C. Sharma, D. P. Acharya and K. Aramaki, Langmuir, 2007, 23, 5324–5330.
43. S. A. Safran and J. Klein, J. Phys. II, 1993, 3, 749–757.
44. J. Jiao and D. J. Burgess, AAPS PharmSci, 2003, 5, 62–73.
45. N. Dixit, K. Maloney and D. Kalonia, Pharm. Res., 2013, 30, 1848–1859.
46. L. J. Peltonen and J. Yliruusi, J. Colloid Interface Sci., 2000, 227, 1–6.
47. J. H. Irving and J. G. Kirkwood, J. Chem. Phys., 1950, 18, 817–829.
48. Y. Li, Y. Guo, M. Bao and X. Gao, J. Colloid Interface Sci., 2011, 361, 573–580.
49. Q. Zhou and M. J. Rosen, Langmuir, 2003, 19, 4555–4562.
50. P. Posocco, M. Fermeglia and S. Pricl, J. Mater. Chem., 2010, 20, 7742–7753.
51. H. Goel, P. R. Chandran, K. Mitra, S. Majumdar and P. Ray, Estimation of interfacial tension for immiscible and partially miscible liquid systems by Dissipative Particle Dynamics, Chem. Phys. Lett., 2014, 600, 62–67.
52. G. Scocchi, P. Posocco, J.-W. Handgraaf, J. G. E. M. Fraaije, M. Fermeglia and S. Pricl, Chem.–Eur. J., 2009, 15, 7586–7592.
53. G. Scocchi, P. Posocco, M. Fermeglia and S. Pricl, J. Phys. Chem. B, 2007, 111, 2143–2151.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra24262b

---